checkpoint

Author: HeidiSteen

articles/search/TOC.yml

@@ -13,11 +13,11 @@
     href: search-faq-frequently-asked-questions.yml
 - name: Quickstarts
   items:
-  - name: Vector search
+  - name: Vector store
     href: search-get-started-vector.md
   - name: Full text search
     href: search-get-started-text.md
-  - name: Semantic ranking
+  - name: Semantic reranking
     href: search-get-started-semantic.md
   - name: Retrieval Augmented Generation (RAG)
     href: search-get-started-retrieval-augmented-generation.md
@@ -129,10 +129,10 @@
   items:
   - name: Storage (indexes)
     items:
-    - name: Vector store
-      href: vector-search-overview.md
     - name: Search index
       href: search-what-is-an-index.md
+    - name: Vector store
+      href: vector-search-overview.md
     - name: Knowledge store 
       href: knowledge-store-concept-intro.md
     - name: Data import strategies
@@ -254,7 +254,7 @@
         href: search-synonyms-tutorial-sdk.md
   - name: Vector stores
     items:
-    - name: Create a vector index
+    - name: Create a vector store
       href: vector-search-how-to-create-index.md
     - name: Configure a vectorizer (preview)
       href: vector-search-how-to-configure-vectorizer.md

articles/search/hybrid-search-overview.md

@@ -9,7 +9,7 @@ ms.service: cognitive-search
 ms.custom:
   - ignite-2023
 ms.topic: conceptual
-ms.date: 11/01/2023
+ms.date: 01/29/2024
 ---
 
 # Hybrid search using vectors and full text in Azure AI Search
@@ -24,9 +24,9 @@ This article explains the concepts, benefits, and limitations of hybrid search.
 
 ## How does hybrid search work?
 
-In Azure AI Search, vector indexes containing embeddings can live alongside textual and numerical fields allowing you to issue hybrid full text and vector queries. Hybrid queries can take advantage of existing functionality like filtering, faceting, sorting, scoring profiles, and [semantic ranking](semantic-search-overview.md) in a single search request.
+In Azure AI Search, vector fields containing embeddings can live alongside textual and numerical fields, allowing you to formulate hybrid queries that execute in parallel. Hybrid queries can take advantage of existing functionality like filtering, faceting, sorting, scoring profiles, and [semantic ranking](semantic-search-overview.md) in a single search request.
 
-Hybrid search combines results from both full text and vector queries, which use different ranking functions such as BM25 and HNSW. A [Reciprocal Rank Fusion (RRF)](hybrid-search-ranking.md) algorithm is used to merge results. The query response provides just one result set, using RRF to determine which matches are included.
+Hybrid search combines results from both full text and vector queries, which use different ranking functions such as BM25 and HNSW. A [Reciprocal Rank Fusion (RRF)](hybrid-search-ranking.md) algorithm merges the results. The query response provides just one result set, using RRF to pick the most relevant matches from each query.
 
 ## Structure of a hybrid query
 

articles/search/index.yml

@@ -64,7 +64,7 @@ landingContent:
     linkLists:
       - linkListType: how-to-guide
         links:
-          - text: Create a vector index in AI Studio
+          - text: Create a vector store in AI Studio
             url: /azure/ai-studio/how-to/index-add
           - text: Build a question and answer copilot
             url: /azure/ai-studio/tutorials/deploy-copilot-ai-studio

articles/search/media/langchain-example-tiktoken-histogram.png renamed to articles/search/media/vector-search-how-to-chunk-documents/langchain-example-tiktoken-histogram.png

@@ -104,7 +104,7 @@ sections:
       - question: |
           How does vector search work in Azure AI Search?
         answer: |
-          With standalone vector search, you first use a deep neural network (DNN), such as a large language model (LLM), to transform content into a vector representation within an embedding space. You can then provide these vectors in a document payload to the search index for indexing. To serve search requests, you use the same DNN from indexing to transform the search query into a vector representation, and vector search finds the most similar vectors and return the corresponding documents.
+          With standalone vector search, you first use an embedding model to transform content into a vector representation within an embedding space. You can then provide these vectors in a document payload to the search index for indexing. To serve search requests, you use the same DNN from indexing to transform the search query into a vector representation, and vector search finds the most similar vectors and return the corresponding documents.
           
           In Azure AI Search, you can index vector data as fields in documents alongside textual and other types of content. The data type for a vector field is `Collection(Edm.Single)`. 
           

articles/search/search-faq-frequently-asked-questions.yml

@@ -178,7 +178,7 @@ But you'll also want to become familiar with methodologies for loading an index
 
 + [Create a search index](search-how-to-create-search-index.md)
 
-+ [Create a vector index](vector-search-how-to-create-index.md)
++ [Create a vector store](vector-search-how-to-create-index.md)
 
 + [Create an index alias](search-how-to-alias.md)
 

articles/search/search-what-is-an-index.md

@@ -35,7 +35,7 @@ You can use the [**Import and vectorize data wizard**](search-get-started-portal
 
 + A skillset that performs data chunking and vectorization of those chunks. You can omit a skillset if you only want integrated vectorization at query time, or if you don't need chunking or [index projections](index-projections-concept-intro.md) during indexing. This article assumes you already know how to [create a skillset](cognitive-search-defining-skillset.md).
 
-+ An index that specifies vector and non-vector fields. This article assumes you already know how to [create a vector index](vector-search-how-to-create-index.md) and covers just the steps for adding vectorizers and field assignments.
++ An index that specifies vector and non-vector fields. This article assumes you already know how to [create a vector store](vector-search-how-to-create-index.md) and covers just the steps for adding vectorizers and field assignments.
 
 + An [indexer](search-howto-create-indexers.md) that drives the pipeline.
 

articles/search/vector-search-how-to-configure-vectorizer.md

@@ -65,7 +65,7 @@ Data chunking (Text Split skill) is free and available on all Azure AI services
 
 + Subdivide large documents into chunks, useful for vector and non-vector scenarios. For vectors, chunks help you meet the input constraints of embedding models. For non-vector scenarios, you might have a chat-style search app where GPT is assembling responses from indexed chunks. You can use vectorized or non-vectorized chunks for chat-style search.
 
-+ Build a vector store where all of the fields are vector fields, and the document ID (required for a search index) is the only string field. Query the vector index to retrieve document IDs, and then send the document's vector fields to another model.
++ Build a vector store where all of the fields are vector fields, and the document ID (required for a search index) is the only string field. Query the vector store to retrieve document IDs, and then send the document's vector fields to another model.
 
 + Combine vector and text fields for hybrid search, with or without semantic ranking. Integrated vectorization simplifies all of the [scenarios supported by vector search](vector-search-overview.md#what-scenarios-can-vector-search-support).
 

articles/search/vector-search-integrated-vectorization.md

@@ -130,7 +130,7 @@ Azure AI Search currently supports the following algorithms:
 
 Within an index definition, you can specify one or more algorithms, and then for each vector field specify which algorithm to use:
 
-+ [Create a vector index](vector-search-how-to-create-index.md) to specify an algorithm in the index and on fields.
++ [Create a vector store](vector-search-how-to-create-index.md) to specify an algorithm in the index and on fields.
 
 + For exhaustive KNN, use [2023-11-01](/rest/api/searchservice/indexes/create-or-update), [2023-10-01-Preview](/rest/api/searchservice/indexes/create-or-update?view=rest-searchservice-2023-10-01-preview&preserve-view=true), or Azure SDK beta libraries that target either REST API version.
 
@@ -154,6 +154,6 @@ Azure AI Search uses HNSW for its ANN algorithm.
 ## Next steps
 
 + [Try the quickstart](search-get-started-vector.md)
-+ [Learn more about vector indexing](vector-search-how-to-create-index.md)
++ [Learn more about vector stores](vector-search-how-to-create-index.md)
 + [Learn more about vector queries](vector-search-how-to-query.md)
 + [Azure Cognitive Search and LangChain: A Seamless Integration for Enhanced Vector Search Capabilities](https://techcommunity.microsoft.com/t5/azure-ai-services-blog/azure-cognitive-search-and-langchain-a-seamless-integration-for/ba-p/3901448)

articles/search/vector-search-overview.md

@@ -40,11 +40,11 @@ HNSW is recommended for most scenarios due to its efficiency when searching over
 
 ## How nearest neighbor search works
 
-Vector queries execute against an embedding space consisting of vectors generated from the same embedding model. Generally, the input value within a query request is fed into the same machine learning model that generated embeddings in the vector index. The output is a vector in the same embedding space. Since similar vectors are clustered close together, finding matches is equivalent to finding the vectors that are closest to the query vector, and returning the associated documents as the search result.
+Vector queries execute against an embedding space consisting of vectors generated from the same embedding model. Generally, the input value within a query request is fed into the same machine learning model that generated embeddings in the vector store. The output is a vector in the same embedding space. Since similar vectors are clustered close together, finding matches is equivalent to finding the vectors that are closest to the query vector, and returning the associated documents as the search result.
 
 For example, if a query request is about hotels, the model maps the query into a vector that exists somewhere in the cluster of vectors representing documents about hotels. Identifying which vectors are the most similar to the query, based on a similarity metric, determines which documents are the most relevant.
 
-When vector fields are indexed for exhaustive KNN, the query executes against "all neighbors". For fields indexed for HNSW, the search engine uses an HNSW graph to search over a subset of nodes within the vector index.
+When vector fields are indexed for exhaustive KNN, the query executes against "all neighbors". For fields indexed for HNSW, the search engine uses an HNSW graph to search over a subset of nodes within the vector store.
 
 ### Creating the HNSW graph